Global Wire

Go back

Ethnic diversity matters for decentralisation and development

Published 03 Aug 2018

Twenty years have now passed since the New Order regime was overthrown in Indonesia. This event triggered not only democratisation in Indonesia but also a remarkable experiment with decentralisation that saw significant power transferred from Jakarta down to the country’s many and diverse districts. While decentralisation likely played an important role in mitigating the centripetal tendencies that gripped Indonesia at the time, there is little evidence that it has effectively delivered on its many other promises, like promoting growth or improving governance.

A father walks his child to school near Sekolah Darurat Kartini, Jakarta, Indonesia. (Photo: Reuters/Beawiharta).

This is consistent with the experiences of regional neighbours that wrangled with decentralisation at the same time, such as the Philippines and Vietnam. Now as Malaysia eyes decentralisation as an important part of its post-transition reform process and Myanmar considers similar reforms as part of its peace process, it is worth revisiting the complexities of devolving power in ethnically diverse settings.

Evidence from Indonesia provides potentially important guidance. It has long been understood that ethnic diversity, all else equal, is associated with poorer provision of public goods including schools, roads and health clinics. This has obvious implications for Southeast Asia, with its vast ethnic, religious and linguistic diversity. But considering ethnic diversity only at highly aggregated levels may be insufficient or outright misleading: the spatial distribution of ethnic diversity — essentially how segregated or intermingled groups are at the local level — almost certainly has implications for its effects.

New research shows that patterns of ethnic segregation significantly affect the relationship between diversity and public goods provision. Locally segregated communities have substantial advantages in procuring public goods relative to more mixed communities, even after taking into account the overall levels of diversity and other potential factors.

This spatial dimension has hitherto not received a lot of attention. But its logic is quite straightforward. Procuring public goods requires coordination between the level where public goods are used — typically a village or other small community — and the level where proposals are approved and funded. In the Indonesian context, proposals for many goods are approved at the district level.

Following previous theories, it is clear that locally segregated communities are better able to coordinate their lobbying efforts than diverse communities. Indeed, there is a virtuous cycle of sorts that emerges when several locally segregated communities exist in relative proximity. This is because one community can point to provisions in a nearby community when making demands for more resources. In short, a sibling rivalry-like effect emerges that incrementally ratchets up public goods in a given area. This dynamic is absent in areas where ethnic groups are more mixed.

This is evident in population census and village census data from Indonesia. What it means in practical terms is that locally segregated areas have a higher provision of public goods than areas that are locally mixed, at least for those goods that are decided upon at the district level. In some instances, diverse areas that are locally segregated have a greater provision of public goods than areas with little ethnic diversity. This has strong implications for effective development, given that public goods form the foundation for sustained growth.

What does this have to do with decentralisation? For decentralisation to support more robust economic development, districts must be able to compete effectively with one another under conditions of greater autonomy. It is generally understood that when there are significant disparities in human, infrastructural or natural resources between areas, then well-endowed areas will outcompete those with fewer resources at their disposal. This potentially widens — rather than shrinks — the developmental gap. Since locally mixed areas may be at a distinct disadvantage when compared to locally segregated areas, their ability to compete effectively may also be inhibited.

The policy implication is clear. Decentralisation provides local areas with greater degrees of autonomy. But it also places a greater burden of self-sufficiency on localities at the very moment that they face increased competition from other areas. To ensure that this does not result in some areas being left behind, central governments must offer targeted support to potentially disadvantaged communities and work to level the playing field.

In the past, levels of human and physical capital were the main criteria in assessing community needs. The data from Indonesia makes clear that patterns of local diversity should also be taken into account so as to ensure that locally mixed communities do not fall behind their more segregated counterparts.

Kai Ostwald is Assistant Professor in the School of Public Policy & Global Affairs and the Department of Political Science at the University of British Columbia.

Krislert Samphantharak is Associate Professor and Associate Dean in the School of Global Policy and Strategy at the University of California San Diego.

Yuhki Tajima is an Assistant Professor in the Edmund A Walsh School of Foreign Service at Georgetown University.

This article is based on the authors’ recent research in the American Political Science Review, ‘Ethnic Segregation and Public Goods: Evidence from Indonesia’.

This article was first published by The East Asia Forum on 25 July 2018. This article was republished with permission.